D Gene Segment 
 
The existence of D genes encoding a short region of the third complementarity determining region )CDR3) of immunoglobulin heavy chains was first postulated when it was apparent that expressed H chains between the sequences encoded by the variable (V) and J (joining) genes contained a segment variable in length that had no counterpart in either gene. It turned out that variability of this region resulted from 2 causes: One was the existence of a set of D genes (D for Diversity); the second was due to the presence of nucleotides that were not germline encoded but added enzymatically at both the VD and the DJ junctions by terminal-deoxynucleotidyl-transferase, which is specifically expressed in lymphocytes. As the result of these combined events, heavy-chain CDR3 appeared as the most hypervariable region of Ig molecules, and therefore of prime importance as a basis for antibody specificity.
 D gene segments appear to be present in all heavy-chain gene loci, from the lowest vertebrate species. In humans, about 30 D gene segments have been identified, clustered between the IgVH locus and the JH genes. The various D genes differ in length, but are always limited to a small number of potential codons. They are flanked on both their 5′ and 3′ ends by the conventional recombination signal sequences recognized by the RAG1 and RAG2 recombinases. The spacers are of different length on each side, so they meet with the 12/23 alternate size rule, to recombine with VH and JH gene segments. D genes make a major contribution to heavy-chain variability of CDR3, not only because of their number and sequence diversity, but also because they are used in many alternate possibilities: (a) They may be inserted in either orientation; (b) any reading frame may be used (in humans, not in the mouse; see text below), although one is more frequently encountered; (c( they may join to each other, so that examples of as many as four fused D genes have been described, thus providing a CDR3 of unusual length; and (d) they may be partially deleted at both ends before being inserted between the V and J genes. All of this clearly can generate a huge diversity potential.
In the mouse, minor differences from the above situation in D gene use and function have been reported, due to the fact that there is essentially only one reading frame used by murine D genes. There are two reasons for this. One gene rearrangement introduces a stop codon in the 3′ JC sequence. The second reason is more complicated. Some mouse D genes have a promoter-like region that can be activated after the first gene rearrangement making the DJ joint has been performed. In that case, a D-J-Cm protein is synthesized, which becomes exposed at the surface of the preB cell and blocks any further gene rearrangement, but is unable to contribute a functional immunoglobulin.
 The genes for T-cell receptors (TCRs) have also D genes on the b and d chains. There are only a number of D genes, which are used in either orientation and reading frame. They have 12-bp spacers of identical length on both 5′ and 3′ flanking recombination signal sequences, corresponding to the 23- bp spacers of the V and J genes. TCR D genes also contribute to CDR3, a region that is most variable, as for immunoglobulins.